The present invention relates to the synthesis of cytotoxic anti-tumor antibiotics such as CC-1065 and analogs thereof. In particular, the present invention provides an improved synthesis for seco(xe2x88x92)CBI (5-hydroxy-3-amino-1-[S]-(chloromethyl)-1,2-dihydro-3H-benz(e)indole), and for the synthesis therefrom of CC-1065 analogs comprising a cyclopropabenzindole (CBI) alkylating moiety, which may be incorporated into cell-targeted therapeutic agents.
CC-1065 is a highly cytotoxic anti-tumor antibiotic isolated from cultures of Streptomyces zelensis. The CC-1065 molecule consists of three substituted pyrroloindole subunits linked by amide bonds. The xe2x80x9cAxe2x80x9d subunit is the alkylating cyclopropapyrroloindole (CPI) moiety, while the xe2x80x9cBxe2x80x9d and xe2x80x9cCxe2x80x9d subunits are identical pyrroloindole moieties. 
Novel cytotoxic agent-cell binding agent conjugates comprising a cell-binding, agent chemically linked to analogs of CC-1065 have been described [U.S. Pat. Nos. 5,475,092; 5,585,499; 5,846,545, R. V. J. Chari et al., Cancer Res., 55, 4079-4084 (1995)]. These cytotoxic agent-cell binding agent conjugates have therapeutic use because they deliver the cylotoxic agent to a specific cell population in a targeted fashion. In these cytotoxic agents, herein called DC1 and its derivatives, the alkylating CPI subunit xe2x80x9cAxe2x80x9d was replaced by the benzannelated analog cyclopropabenzindole (CBI). 
CBI is the precursor required for the synthesis of DC1 drugs and its derivatives. The original synthesis of CBI was described by D. L. Boger et al., [J. Org. Chem., 55, 5823-5833 (1990)]. An xe2x80x9cimprovedxe2x80x9d synthesis, also described by D. L. Boger et al., [J. Org. Chem., 57, 2873-2876 (1992)] is a 15-step process starting from naphthalene diol. Other pathways for the syntheses of CBI from different starting materials have also been described [K. J. Drost and M. P. Cava, J. Org. Chem., 56, 2240-2244 (1991), P. A. Aristoff and P. D. Johnson, J Org. Chem., 57, 6234-6239 (1992)]. These syntheses are lengthy, time-consuming, expensive and provide poor yields.
A key step in the synthesis of CBI is the resolution of the enantiomers at the seco-CBI stage. Only the seco(xe2x88x92)enantiomer is biologically active, and it is important to efficiently remove the inactive (+) isomer. Isomer separation can be achieved, for example, by chiral HPLC. This method is not very efficient when applied to seco-CBI because the separation between the two enantiomers is poor. In addition, even the optimized separation on a chiral column is poor (retention time difference between the two isomers is less than 5 minutes), and requires a very non-polar solvent system, such as a mixture of 95% hexane and 5% isopropanol (Boger et al., 116, J.Am. Chem Soc., 7996-8006 (1994). Under these conditions, seco-CBI is poorly soluble, resulting in low efficiency (small loading amounts) on the column, and thus, long processing times. Alternatively, the enantiomeric mixture can be converted into a set of diastereomers by esterification with a chiral acid, such as mandelic acid, followed by separation by HPLC. However, the separated ester has to be hydrolyzed and then repurified, thus adding an extra processing step.
The therapeutic utility and promise of drugs such as DC1 and its derivatives, for example in the treatment of various cancers, makes it desirable that improved synthetic methods be developed in order to be able to manufacture CBI in large scale, by a simple, easily scalable, high-yield, inexpensive process that uses inexpensive and easily available starting materials.
The present invention provides such an improved synthetic method that addresses the aforementioned shortcomings of the prior art. All these advantages and more are provided by the invention described herein, as will be apparent to one of skill in the art upon reading the following disclosure and examples.
The inventors have discovered a new, economical and efficient synthesis for seco(xe2x88x92)CBI that can utilize, for example, the commercially available and inexpensive compound 1,3-dihydroxynaphthalene as a starting material, and which can be accomplished in as few as seven steps.
The inventors have further provided related flexible and efficient syntheses for the conversion of seco(xe2x88x92)CBI into a wide variety of DC1 drugs. While there are several differences between the synthetic scheme for seco(xe2x88x92)CBI described herein and any previously reported method, one exemplary difference is the use of the same protecting group for the amino and the hydroxy groups of the key precursor, 4-hydroxy-2-naphthyl amine. Thus, in one embodiment of the method described herein, a di-tert.-butyloxycarbonyl (di-t-boc) protected compound is used, instead of a separate benzyl protecting group for the hydroxyl group and a tert.-butyloxycarbonyl (t-boc) protecting group for the amine function, described previously. Thus, in the present syntheses, some of the redundant protection and deprotection steps have been removed. These and other changes have shortened the synthesis time, improved the product yield considerably, and also improved the separation of enantiomers.
In the present invention, the use of two t-boc protecting groups is preferred and gives a seco-CBI enantiomeric mixture that separates well on a chiral HPLC column. In addition, the column can be run with a solvent mixture with a higher polarity, for example containing 20% isopropanol, in which the compound has good solubility. These two features greatly increase the loading capacity of the column and therefore the efficiency of the separation process, and thus decrease the processing time considerably.
Thus, in a first aspect, the present invention provides a process for preparing the seco(xe2x88x92)CBI of formula (I): 
in which a di-protected compound of formula (II) is used, in which R is a protecting group such that the amino group and hydroxyl group are protected by the same compound: 
and the compound of formula (II) is converted by alkylation and ring-closure reactions to provide a racemic mixture represented by a compound of formula (III): 
The (xe2x88x92) isomer of racemate (III) can be isolated, for example by chiral chromatography, and the isolated (xe2x88x92) isomer of the compound of formula (III) is deprotected to produce the compound of formula (I).
In preferred embodiments, R is tert-butyloxycarbonyl and the alkylation step employs 1,3-dichloropropene.
In certain embodiments, a compound of formula (II) can be conveniently prepared from an inexpensive and easily obtained starting material such as 1,3-dihydroxynapthalene by amination and protection of the hydroxyl and amine groups (FIG. 1).
In a second aspect, the present invention provides a process for preparing DC1 by reacting the amino group of a compound of the seco(xe2x88x92)CBI of formula (I) to form a peptide bond, where the seco(xe2x88x92)CBI may be prepared according to the method of the present invention.
Thus, in a first embodiment of this second aspect of the invention, a peptide bond is formed by reacting the amino group of seco(xe2x88x92)CBI with the carboxyl group of, for example, a compound of formula (IV) under suitable conditions, 
in which R1 represents, in this embodiment, an alkyl or aryl thio group that forms a disulfide bond within a compound of formula (IV), such as, for example, an alkyl or aryl thiol, or, more specifically, xe2x80x94Sxe2x80x94CH3 or xe2x80x94S-pyridyl. Such disulfides can be used to link the DC1 compound to, for example, a cell-targeting agent via a bond that can be cleaved inside the target cell.
This embodiment is not limited to only the synthesis of the DC1 compound corresponding to the product of the reaction using compound (IV), but can also be readily adapted to produce a wide variety of DC1 compounds, including those in which the group that is capable of bonding to a cell-targeting agent can be other than a thio or disulfide group, such as, for example, an acid-labile group, a photo-labile group, a peptidase-labile group, or an esterase-labile group, depending upon the analog of compound (IV) that is selected.
In a further embodiment of this aspect of the invention, it is not required that the coupling of seco(xe2x88x92)CBI occurs as a terminal step of the synthesis. Thus, in this embodiment the DC1 compound is synthesized from a bis-indolyl moiety, a disulfide-containing moiety, and seco(xe2x88x92)CBI, by their attachment via peptide bonds, and the order in which these three components of DC1 are assembled is not critical. For example, the bis-indolyl moiety and seco(xe2x88x92)CBI can be linked prior to the attachment of the disulfide-containing moiety.